Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Renkai/rikai-example

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

41 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Rikai Examples

Convention

Use /tmp/rikai_example as rikai warehouse, for xyz.py, save rikai formatted data into /tmp/rikai_example/{xyz}[optional].

Getting Started

$ conda create --no-default-packages -n rikai-example python=3.8 --yes
$ conda activate rikai-example
$ pip install -r requirements.txt
$ python coco_sample.py
$ python
>>> from example import spark
>>> df = spark.read.format("rikai").load("/tmp/rikai_example/coco")
>>> df.show(1, truncate=False, vertical=True)
-RECORD 0---------------------------------------------------------------------------------------------------------------------------------------------------------------------
 image       | Image(uri='Some(coco_sample/train_sample/000000070211.jpg)')
 image_id    | 70211
 annotations | [{Box2d(xmin=135.0, ymin=13.05, xmax=230.55, ymax=36.379999999999995), book, 84}, {Box2d(xmin=325.53, ymin=0.0, xmax=362.53999999999996, ymax=40.62), tv, 72}]
only showing top 1 row

Table of contents

Code Description Output
coco_sample.py prepare the coco_sample dataset /tmp/rikai_example/coco_sample, /tmp/rikai_example/coco_labels
resnet_yaml.py create resnet model via yaml
resnet_mlflow.py create resnet model via mlflow
yolov5_mlflow.py create yolov5 model via mlflow
yolov5_video.py prepare the dataset for the yolov5_video notebook /tmp/rikai_example/yolov5_video
sql_show_models.py Example of show models
coco_sample.ipynb explore the coco_sample dataset in JupyterLab
yolov5_mlflow.ipynb Applying yolov5 on the coco_sample dataset
yolov5_video.ipynb Applying yolov5 on the Elephant Dream Video

About

**Evolving, not stable!!!** Minimum examples for Rikai.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 99.2%
  • Python 0.8%